Skip to content

cmey/Speckle2Speckle

 
 

Repository files navigation

Speckle2Speckle

This implementation of Speckle2Speckle is based on code of Joey Litalien https://github.com/joeylitalien/noise2noise-pytorch

License

The files in the folder src can be used under the MIT license (see the LICENSE file there).

The implementations of the comparison methods (evaluation/otherMethods/SRAD/ and evaluation/otherMethods/OBNLMpackage/) follow their own licensing.

All other content including the images can only be used for research purposes.

Unpack Zips

Unpack the following ZIPs in their folders:

  • data/test-set_simulated.zip: It contains the test-set of the simulated data.
  • outputImages/otherMedhods_simulated.zip: Contains the precomputed output of the other methods on the simulated test-set

Apply Network to test-set, phantom and invivo

runInference.sh

Apply Other Methods

Precomputed images are included as the application takes significant time.

evaluation/otherMethods/otherMethods_experimental.m
evaluation/otherMethods/otherMethods_invivo.m
evaluation/otherMethods/otherMethods_simulated.m

Compute Metrics & Create Plots

cd src
python3 compute_metrics.py

About

DL de-speckling

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 69.6%
  • MATLAB 19.8%
  • C 8.6%
  • Shell 2.0%